Autism is a pervasive development disorder characterized by profound deficits in many aspects of social communication. It is now widely accepted that the origins of autistic symptoms are genetic and are manifested in neurobiological dysfunction. Nonetheless, identifying the specific neuroanatomical underpinnings of autistic symptomatology has proved elusive. Most notably, answers to the causes of autistic symptoms are not to be found in macroscopic structural pathology. A promising alternative to structural explanations is offered by functional imaging, which can measure regional brain function as well as functional connectivity. Functional magnetic resonance imaging (fMRI) has recently emerged as a highly sensitive, non-invasive means of determining brain function in both normal and impaired populations. Functional MRI measures changes in deoxyhemoglobin levels resulting from local increases in cerebral blood flow, thus providing an indirect measure of focal brain activity associated with task performance. This project proposes using fMRI in high functioning autistic subjects while performing cognitive tasks that tap three primary deficits associated with autism: these are emotional responsiveness and empathy, perception and identification of facial emotions, and perception and attribution of affective prosody. The tasks are designed to isolate different levels of cognitive function essential to social communication: specifically, the effects of emotional arousal and responsiveness, perception of affective information, and the execution of utilization of affective information. Using newly established statistical techniques, the fMRI project will examine not only which brain areas respond to cognitive challenge in autistic vs. control subjects, but will also define the functional correlations between different cognitive processing regions for each population. This approach is geared toward testing the hypothesis that the primary deficits in autism are not structural, but are reflected in impaired functional connectivity between the analysis of affective information, and the dynamic utilization of that information in problems-solving tasks.